Regression Analysis

Regression analysis is a statistical tool for the investigation of re- lationships

between variables. Usually, the investigator seeks to ascertain the causal effect of one
variable upon anotherthe effect of a price increase upon demand, for example, or the
effect of changes in the money supply upon the inflation rate. To explore such issues, the
investigator assembles data on the underlying variables of interest and employs
regression to estimate the quantitative effect of the causal variables upon the variable
that they influence. The investigator also typically assesses the statistical significance
of the estimated relationships, that is, the degree of confidence that the true
relationship is close to the estimated relationship.
What is regression?
For purposes of illustration, suppose that we wish to identify and quantify the factors
that determine earnings in the labor market. A moments reflection suggests a myriad of
factors that are associated with variations in earnings across individualsoccupation,
age, ex- perience, educational attainment, motivation, and innate ability come to
mind, perhaps along with factors such as race and gender that can be of particular
concern to lawyers. For the time being, let us restrict attention to a single factorcall it
education. Regression analysis with a single explanatory variable is termed simple
regres- sion.
1. Simple Regression
In reality, any effort to quantify the effects of education upon earnings without
careful attention to the other factors that affect earnings could create serious statistical
difficulties (termed omitted variables bias), which I will discuss later. But for now let us
assume away this problem. We also assume, again quite unrealistically, that education
can be measured by a single attributeyears of school- ing. We thus suppress the fact
that a given number of years in school may represent widely varying academic programs.
At the outset of any regression study, one formulates some hy- pothesis about the
relationship between the variables of interest, here, education and earnings. Common
experience suggests that better educated people tend to make more money. It further
suggests that the causal relation likely runs from education to earnings rather than the
other way around. Thus, the tentative hypothesis is that higher levels of education
cause higher levels of earnings, other things being equal.
To investigate this hypothesis, imagine that we gather data on education and
earnings for various individuals. Let E denote educa- tion in years of schooling for each
individual, and let I denote that individuals earnings in dollars per year. We can plot
this informa- tion for all of the individuals in the sample using a two-dimensional
diagram, conventionally termed a scatter diagram. Each point in the diagram
represents an individual in the sample.
The diagram indeed suggests that higher values of E tend to yield higher values of
I, but the relationship is not perfectit seems that knowledge of E does not suffice for
an entirely accurate predic- tion about I.
8
We can then deduce either that the effect of
education upon earnings differs across individuals, or that factors other than education
influence earnings. Regression analysis ordinarily embraces the latter explanation.
q
Thus, pending discussion below of omitted variables bias, we now hypothesize that
earnings for each individual are determined by education and by an aggregation of
omitted factors that we term noise.
To refi ne the hypothesis further, it is natural to suppose that people in the labor
force with no education nevertheless make some
positive amount of money, and that education increases earnings above this baseline.
We might also suppose that education affects in- come in a linear fashionthat is, each
additional year of schooling adds the same amount to income. This linearity assumption
is com- mon in regression studies but is by no means essential to the appli- cation of the
technique, and can be relaxed where the investigator has reason to suppose a priori
that the relationship in question is nonlinear.
1o
Then, the hypothesized relationship between education and earnings may be
written
where
I = + E +
= a constant amount (what one earns with zero education);
= the effect in dollars of an additional year of schooling on in- come, hypothesized to be
positive; and
= the noise term reflecting other factors that influence earn- ings.
The variable I is termed the dependent or endogenous vari- able; E is termed
the independent, explanatory, or exogenous variable; is the constant term and
the coefficient of the vari- able E.
Remember what is observable and what is not. The data set contains observations
for I and E. The noise component is com- prised of factors that are unobservable, or
at least unobserved. The parameters and are also unobservable. The task of
regression analysis is to produce an estimate of these two parameters, based
upon the information contained in the data set and, as shall be seen, upon some
assumptions about the characteristics of .
To understand how the parameter estimates are generated, note that if we ignore the
noise term , the equation above for the rela- tionship between I and E is the equation
for a linea line with an intercept of on the vertical axis and a slope of .
Returning to the scatter diagram, the hypothesized relationship thus implies that
somewhere on the diagram may be found a line with the equation I
= + E. The task of estimating and is equivalent to the task of estimating where this
line is located.
What is the best estimate regarding the location of this line? The answer depends in
part upon what we think about the nature of the noise term . If we believed that was
usually a large negative num- ber, for example, we would want to pick a line lying above
most or all of our data pointsthe logic is that if is negative, the true value
of I (which we observe), given by I = + E + , will be less than the value of I on the
line I = + E. Likewise, if we believed that was systematically positive, a line lying
below the majority of data points
would be appropriate. Regression analysis assumes, however, that
the noise term has no such systematic property, but is on average equal to zeroI will
make the assumptions about the noise term more precise in a moment. The
assumption that the noise term is usually zero suggests an estimate of the line that lies
roughly in the midst of the data, some observations below and some observations
above.
But there are many such lines, and it remains to pick one line in particular.
Regression analysis does so by embracing a criterion that relates to the estimated noise
term or error for each observation. To be precise, define the estimated error for each
observation as the
vertical distance between the value of I along the estimated line I =
+ E (generated by plugging the actual value of E into this equation) and the true value
of I for the same observation. Superimposing a candidate line on the scatter diagram,
the estimated errors for each observation may be seen as follows:
With each possible line that might be superimposed upon the data, a different set of
estimated errors will result. Regression analysis then chooses among all possible lines by
selecting the one for which the sum of the squares of the estimated errors is at a
minimum. This is termed the minimum sum of squared errors (minimum SSE) crite-
rion The intercept of the line chosen by this criterion provides the
estimate of , and its slope provides the estimate of .
It is hardly obvious why we should choose our line using the minimum SSE
criterion. We can readily imagine other criteria that might be utilized (minimizing the
sum of errors in absolute value,
11
for example). One virtue of the SSE criterion is that it is
very easy to employ computationally. When one expresses the sum of squared errors
mathematically and employs calculus techniques to ascertain
the values of and that minimize it, one obtains expressions for
and that are easy to evaluate with a computer using only the observed values of E
and I in the data sample.
1:
But computational convenience is not the only virtue of the
minimum SSE criterionit also has some attractive statistical properties under
plausible as- sumptions about the noise term. These properties will be discussed in a
moment, after we introduce the concept of multiple regression.
r. Multiple Regression
Plainly, earnings are affected by a variety of factors in addition to years of schooling,
factors that were aggregated into the noise term in the simple regression model above.
Multiple regression is a technique that allows additional factors to enter the analysis
sepa- rately so that the effect of each can be estimated. It is valuable for quantifying
the impact of various simultaneous influences upon a single dependent variable.
Further, because of omitted variables bias with simple regression, multiple regression is
often essential even when the investigator is only interested in the effects of one of the
independent variables.
For purposes of illustration, consider the introduction into the earnings analysis of
a second independent variable called experi- ence. Holding constant the level of
education, we would expect someone who has been working for a longer time to earn
more. Let X denote years of experience in the labor force and, as in the case of
education, we will assume that it has a linear effect upon earnings that is stable across
individuals. The modified model may be written:
I = + E + X +
where is expected to be positive.
The task of estimating the parameters , , and is conceptually identical to the
earlier task of estimating only and . The di fference is that we can no longer think
of regression as choosing a line in a two-dimensional diagramwith two explanatory
variables we need three dimensions, and instead of estimating a line we are
estimating a plane. Multiple regression analysis will select a plane so that the sum of
squared errorsthe error here being the vertical distance between the actual value of I
and the estimated planeis at a minimum. The intercept of that plane with the I-axis
(where E and X are zero) implies the constant term , its slope in the education
dimension implies the coefficient , and its slope in the experience dimension implies
the coefficient .
Multiple regression analysis is in fact capable of dealing with an arbitrarily large
number of explanatory variables. Though people lack the capacity to visualize in more than
three dimensions, mathematics does not. With n explanatory variables, multiple
regression analysis will estimate the equation of a hyperplane in n-space such that the
sum of squared errors has been minimized. Its intercept implies the constant term, and
its slope in each dimension implies one of the regression coefficients. As in the case of
simple regression, the SSE criterion is quite convenient computationally. Formulae for
the parameters , , . . . can be derived readily and evaluated easily on a computer,
again using only the observed values of the dependent and independent variables.
The interpretation of the coefficient estimates in a multiple regression warrants brief
comment. In the model I = + E + X + , captures what an individual earns with no
education or experience, captures the effect on income of a year of education, and
captures the effect on income of a year of experience. To put it slightly differently, is an
estimate of the effect of a year of education on income holdiing experience constant.
Likewise, is the estimated effect of a year of experience on income, holding education
constant.
. Essential Assumptions and Statistical Properties of Regression
As noted, the use of the minimum SSE criterion may be de- fended on two
grounds: its computational convenience, and its desir- able statistical properties. We now
consider these properties and the assumptions that are necessary to ensure them.
1|
Continuing with our illustration, the hypothesis is that earnings in the real world
are determined in accordance with the equation I = + E + X + true values of ,
, and exist, and we desire
to ascertain what they are. Because of the noise term , however, we can only estimate
these parameters.
We can think of the noise term as a random variable, drawn by nature from some
probability distributionpeople obtain an educa- tion and accumulate work experience,
then nature generates a ran- dom number for each individual, called , which
increases or de- creases income accordingly. Once we think of the noise term as a
random variable, it becomes clear that the estimates of , , and (as distinguished from
their true values) will also be random variables, because the estimates generated by the
SSE criterion will depend upon the particular value of drawn by nature for each
individual in the data set. Likewise, because there exists a probability distribution from
which each is drawn, there must also exist a probability dis- tribution from which each
parameter estimate is drawn, the latter distribution a function of the former
distributions. The attractive statistical properties of regression all concern the
relationship be- tween the probability distribution of the parameter estimates and the true
values of those parameters.
We begin with some definitions. The minimum SSE criterion is termed an estimator.
Alternative criteria for generating parameter es- timates (such as minimizing the sum of
errors in absolute value) are also estimators.
Each parameter estimate that an estimator produces, as noted, can be viewed as a
random variable drawn from some probability distribution. If the mean of that
probability distribution is equal to the true value of the parameter that we are trying to
estimate, then the estimator is unbiased. In other words, to return to our illustra- tion,
imagine creating a sequence of data sets each containing the same individuals with the
same values of education and experience,
differing only in that nature draws a different for each individual for each data set.
Imagine further that we recompute our parameter estimates for each data set, thus
generating a range of estimates for each parameter , , and . If the estimator is
unbiased, we would find that on average we recovered the true value of each parameter.
An estimator is termed consistent if it takes advantage of addi- tional data to
generate more accurate estimates. More precisely, a consistent estimator yields estimates
that converge on the true value of the underlying parameter as the sample size gets larger
and larger. Thus, the probability distribution of the estimate for any parameter has lower
variance
1
as the sample size increases, and in the limit (infinite sample size) the
estimate will equal the true value.
The variance of an estimator for a given sample size is also of in- terest. In particular,
let us restrict attention to estimators that are un- biased. Then, lower variance in the
probability distribution of the estimator is clearly desirable

it reduces the probability
of an es- timate that differs greatly from the true value of the underlying pa- rameter. In
comparing different unbiased estimators, the one with the lowest variance is termed
efficient or best.
Under certain assumptions, the minimum SSE criterion has the characteristics of
unbiasedness, consistency, and efficiencythese assumptions and their consequences
follow:
If the noise term for each observation, , is drawn from a distribution that has a
mean of zero, then the sum of squared errors criterion generates estimates that are
unbiased and consistent.
That is, we can imagine that for each observation in the sample, nature draws a
noise term from a different probability distribution. As long as each of these distributions
has a mean of zero (even if the dsitributions are not the same), the minimum SSE
criterion is unbi- ased and consistent.
1 ,
This assumption is logically sufficient to en-
sure that one other condition holdsnamely, that each of the ex- planatory variables
in the model is uncorrelated with the expected value of the noise term.

This will prove
important later.
If the distributions from which the noise terms are drawn for each observation have
the same variance, and the noise terms are statistically independent of each other (so
that if there is a positive noise term for one observation, for example, there is no reason
to ex- pect a positive or negative noise term for any other observation), then the sum
of squared errors criterion gives us the best or most efficient estimates available from
any linear estimator (defined as an estimator that computes the parameter estimates as a
linear function of the noise term, which the SSE criterion does).
1q
If assumptions are violated, the SSE criterion remains unbi- ased and consistent but
it is possible to reduce the variance of the es- timator by taking account of what we
know about the noise term. For example, if we know that the variance of the
distribution from which the noise term is drawn is bigger for certain observations, then the
size of the noise term for those observations is likely to be larger. And, because the noise
is larger, we will want to give those observa- tions less weight in our analysis. The
statistical procedure for dealing
with this sort of problem is termed generalized least squares, which is beyond the scope
of this lecture.
:o
. An IllustrationDiscrimination on the Basis of Gender
To illustrate the ideas to this point, as well as to suggest how re- gression may have
useful applications in a legal proceeding, imagine a hypothetical firm that has been sued
for wage discrimination on the basis of gender. To investigate these allegations, data
have been gathered for all of the firms employees. The questions to be an- swered are
(a) whether discrimination is occurring (liability), and (b) what its consequences are
(damages). We will address them using a modified version of the earnings model
developed in section 1.
The usefulness of multiple regression here should be intuitively apparent. Suppose,
for example, that according to the data, women at the firm on average make less than
men. Is this fact sufficient to establish actionable discrimination? The answer is no if
the differ- ence arises because women at this firm are less well-educated, for ex- ample
(and thus by inference less productive), or because they are less experienced.
: 1
In
short, the legal question is whether women earn less after all of the factors that the
firm may permissibly con- sider in setting wages have been taken into account.
To generate the data for this illustration, I assume a hypothetical
real world in which earnings are determined by equation:
Earnings = 5000 + 1000 School + 50 Aptitude
+ 300 Experience 2000 Gendum + Noise
where School is years of schooling; Aptitude is a score between
100 and 240 on an aptitude test; Experience is years of experience in the work force;
and Gendum is a variable that equals 1 for women and zero for men (more about
this variable in a moment). To produce the artificial data set, I made up fi fty
observations (corresponding to fifty fictitious individuals) for each of the explana- tory
variables, half men and half women. In making up the data, I deliberately tried to
introduce some positive correlation between the
schooling and aptitude variables, for reasons that will become clear later. I then
employed a random number generator to produce a noise term drawn from a normal
distribution, with a standard devia- tion (the square root of the variance) equal to , ooo
and a mean of zero. This standard deviation was chosen more or less arbitrarily to
introduce a considerable but not overwhelming amount of noise in proportion to the
total variation in earnings. The right-hand-side variables were then used to generate the
actual value of earnings for each of the fifty individuals.
The effect of gender on earnings in this hypothetical firm enters through the variable
Gendum. Gendum is a dummy variable in econometric parlance because its numerical
value is arbitrary, and it simply captures some nonnumerical attribute of the sample
popula- tion. By construction here, men and women both earn the same re- turns to
education, experience, and aptitude, but holding these fac- tors constant the earnings of
women are $ , : ooo lower. In effect, the constant term (baseline earnings) is lower for
women, but otherwise women are treated equally. In reality, of course, gender
discrimina- tion could arise in other ways (such as lower returns to education and
experience for women, for example), and I assume that it takes this form only for
purposes of illustration.
Note that the random number generator that I employed here generates noise
terms with an expected value of zero, each drawn from a distribution with the same
variance. Further, the noise terms for the various observations are statistically independent
(the realized value of the noise term for each observation has no influence on the noise
term drawn for any other observation). Hence, the noise terms satisfy the assumptions
necessary to ensure that the minimum SSE criterion yields unbiased, consistent, and
efficient estimates. The ex- pected value of the estimate for each parameter is equal to
the true value, therefore, and no other linear estimator will do a better job at recovering
the true parameters than the minimum SSE criterion. It is nevertheless interesting to
see just how well regression analysis performs. I used a standard computer package to
estimate the con- stant term and the coefficients of the four independent variables
from the observed values of Earnings, School, Aptitude, Experience, and
Gendum for each of the fifty hypothetical individu- als. The results are reproduced in
table 1, under the column labeled
Estimated Value. (We will discuss the last three columns and the
R
2
statistic in the next section.)
Table 1 (noise term with standard deviation of , ooo)
Variable True
value
Estimated S
value
tandard t- Prob
error statistic (:-tail)
Constant ooo.o |16., ,81.8 1.oq| .:8o
School 1ooo.o 18|.6 :88.1 .oo .ooo
Aptitude o.o 6.| :,. o.:6 .81|
Experience oo.o :|1., 8o.8 :.qq: .oo|
Gendum

:ooo.o

1|,o.| 1|o:.:

1.o|q .oo
R
2
= .6|6
Note that all of the estimated parameters have the right sign. Just by chance, it
turns out that the regression overestimates the returns to schooling and underestimates
the other parameters. The estimated coefficient for Aptitude is off by a great deal in
proportion to its true value, and in a later section I will offer an hypothesis as to what the
problem is. The other parameter estimates, though obvi- ously different from the true
value of the underlying parameter, are much closer to the mark. With particular reference
to the coefficient of Gendum, the regression results correctly suggest the presence of
gender discrimination, though its magnitude is underestimated by about : percent
(remember that an overestimate of the same magni- tude was just as likely ex ante, that is,
before the actual values for the noise terms were generated).
The source of the error in the coefficient estimates, of course, is the presence of
noise. If the noise term were equal to zero for every observation, the true values of the
underlying parameters could be recovered in this illustration with perfect accuracy from
the data for only five hypothetical individualsit would be a simple matter of solving
five equations in five unknowns. And, if noise is the source of error in the parameter
estimates, intuition suggests that the magnitude of the noise will affect the accuracy
of the regression estimates, with more noise leading to less accuracy on average. We
will make this intuition precise in the next section, but before
proceeding it is perhaps useful to repeat the parameter estimation experiment for a
hypothetical firm in which the data contain less noise. To do so, I took the data for
the independent variables used in the experiment above and again generated values for
earnings for the fifty hypothetical individuals using equation (1), changing only the
noise terms. This time, the noise terms were drawn by the random number generator
from a normal distribution with standard deviation of , 1 ooo rather than , ooo (a signifi
cant reduction in the amount of noise). Reestimating the regression parameters from this
modified data set produced the results in table : :
Table : (noise term with standard deviation of , 1 ooo)
Variable True
value
Estimated
value
Standard
error
t-statis- tic Prob
(:-tail)
Constant ooo.o |,8|.: q|.| .o6o .ooo
School 1ooo.o 11|6.: ,:.o 1.q1 .ooo
Aptitude o.o q.1 6.8 .,|1 .ooo
Experience oo.o :8.| :o.: 1|.11 .ooo
Gendum

:ooo.o

186,.6 o.

.:8 .ooo
R
2
= .q6|
Not surprisingly, the estimated parameters here are considerably closer to their true
values. It was not certain that they would be, be- cause after all their expected values are
equal to their true values re- gardless of the amount of noise (the estimator is unbiased).
But on average we would expect greater accuracy, and greater accuracy in- deed
emerges here. Put more formally, the probability distributions of the parameter estimates
have greater variance, the greater the vari- ance of the noise term. The variance of the
noise term thus affects the degree of confidence that we have in the accuracy of
regression estimates.
In real applications, of course, the noise term is unobservable, as is the distribution
from which it is drawn. The variance of the noise term is thus unknown. It can,
however, be estimated using the difference between the predicted values of the
dependent variable for each observation and the actual value (the estimated errors
defined
earlier). This estimate in turn allows the investigator to assess the explanatory power of
the regression analysis and the statistical sig- nificance of its parameter estimates.
. Statistical Inference and Goodness of fit
Recall that the parameter estimates are themselves random vari- ables, dependent
upon the random variables . Thus, each estimate can be thought of as a draw from
some underlying probability distri- bution, the nature of that distribution as yet
unspecified. With a further assumption, however, we can compute the probability distri-
bution of the estimates, and use it to test hypotheses about them.
. Statistical Inference
Most readers are familiar, at least in passing, with a probability distribution called
the normal. Its shape is that of a bell curve, indicating among other things that if a
sample is drawn from the distribution, the most likely values for the observations in the
sample are those close to the mean and least likely values are those farthest
from the mean. If we assume that the noise terms are all drawn from the same normal
distribution, it is possible to show that the parameter estimates have a normal distribution
as well.
::
The variance of this normal distribution, however, depends upon the variance of the
distribution from which the noise terms are drawn. This variance is unknown in
practice and can only be esti- mated using the estimated errors from the regression to
obtain an estimate of the variance of the noise term. The estimated variance of the noise
term in turn can be used to construct an estimate of the
::
See, e.g., E. Hanushek and J. Jackson, supra note 1, at 66

68; J. Johnston, supra note 1, at 1

8. The supposition that the noise terms are normally dis- tributed is often intuitively plausible and
may be loosely justified by appeal to central limit theorems, which hold that the average of a large
number of random variables tends toward a normal distribution even if the individual random vari- ables
that enter into the average are not normally distributed. See, e.g., R. Hogg and A. Craig, Introduction to
Mathematical Statistics 1q:

q (|th ed. 1q,8); W. Feller, An Introduction to Probability Theory and Its
Applications, vol. 1, :|

|8 (d ed. 1q68). Thus, if we think of the noise term as the sum of a large
number of in- dependent, small disturbances, theory affords considerable basis for the supposi- tion that its
distribution is approximately normal.
variance of the normal distribution for each coefficient. The square root of this
estimate is called the standard error of the coefficientcall this measure s.
It is also possible to show
:
that if the parameter estimate, call it
, is normally
distributed with a mean of , then (

)/s has a
Students t distribution. The t-
distribution looks very much like the normal, only it has fatter tails and its mean is
zero. Using this result, suppose we hypothesize that the true value of a parameter in our
regression model is . Call this the null hypothesis. Because the minimum SSE
criterion is an unbiased estimator, we can deduce that our parameter estimate is drawn
from a normal distribution with a mean of if the null hypothesis is true. If we then
subtract from our actual parameter estimate and divide by its standard error, we obtain
a number called the t-statistic, which is drawn from a t- distribution if the null
hypothesis is true. This statistic can be positive or negative as the parameter
estimate from which it is derived is greater or less than the hypothesized true value
of the parameter. Recalling that the t-distribution is much like a normal with mean of
zero, we know that large values of the t-statistic (in absolute value) will be drawn
considerably less frequently than small values of the t-statistic. And, from the
construction of the t-statistic, large values for that statistic arise (in absolute value),
other things being equal, when the parameter estimate on which it is based differs from its
true (hypothesized) value by a great deal.
This insight is turned on its head for hypothesis testing. We have just argued that
a large t-statistic (in absolute value) will arise fairly infrequently if the null hypothesis
is correct. Hence, when a large t-statistic does arise, it will be tempting to conclude
that the null hypothesis is false. The essence of hypothesis testing with a regression
coefficient, then, is to formulate a null hypothesis as to its
true value, and then to decide whether to accept or reject it according to
whether the t-statistic associated with that null hypothesis is large enough that the
plausibility of the null hypothesis is sufficiently in doubt.
:|
:
See sources cited note :: supra.
:|
I limit the discussion here to hypothesis testing regarding the value of a particular parameter. In
fact, other sorts of hypotheses may readily be tested, such
One can be somewhat more precise. We might resolve that the null hypothesis is
implausible if the t-statistic associated with our regression estimate lies so far out in one
tail of its t-distribution that such a value, or one even larger in absolute value, would
arise less than, say, percent of the time if the null hypothesis is correct. Put differently,
we will reject the null hypothesis if the t-statistic falls either in the uppermost tail of the
t-distribution, containing :. per- cent of the draws representing the largest positive
values, or in the lowermost tail, containing :. percent of the draws representing the
largest negative values. This is called a two-tailed test.
Alternatively, we might have a strong prior belief about the true value of a parameter
that would lead us to accept the null hypothesis even if the t-statistic lies far out in one of
the tails of the distribution. Consider the coefficient of the gender dummy in table 1 as
an illus- tration. Suppose the null hypothesis is that the true value of this co- efficient is
zero. Under what circumstances would we reject it? We might find it implausible that
the true value of the coefficient would be positive, reflecting discrimination against men.
Then, even if the estimated coefficient for the gender dummy is positive with a large
positive t-statistic, we would still accept the null hypothesis that its true value is zero.
Only a negative coefficient estimate with a large negative t-statistic would lead us to
conclude that the null hypothesis was false. Where we reject the null hypothesis only if
a t-statistic that is large in absolute value has a particular sign, we are employing
a one-tailed test.
To operationalize either a one- or two-tailed test, it is necessary to compute the exact
probability of a t-statistic as large or larger in absolute value as the one associated with
the parameter estimate at issue. In turn, it is necessary to know exactly how spread out
is the t-distribution from which the estimate has been drawn. A further parameter that
we need to pin down the shape of the t-distribution in this respect is called the
degrees of freedom, defined as the number of observations in the sample less the
number of parameters to be estimated. In the illustrations of tables 1 and :, we have
fifty observations in the sample, and we are estimating parameters, so
as the hypothesis that all parameters in the model are zero, the hypothesis that some subset of the
parameters are zero, and so on.
the t-distribution for any of the parameter estimates has | degrees of freedom. The
fewer the degrees of freedom, the more spread out is the t-distribution and thus the
greater is the probability of drawing large t-statistics. The intuition is that the larger the
sample, the more collapsed is the distribution of any parameter estimate (recall the
concept of consistency above). By contrast, the more parameters we seek to estimate
from a sample of given size, the more information we are trying to extract from the
data and the less confident we can be in the estimate of each parameterhence, the
associated t-distribution is more spread out.
:
Knowing the degrees of freedom for the t-distribution allows an investigator to
compute the probability of drawing the t-statistic in question, or one larger in absolute
value, assuming the truth of the null hypothesis. Using the appropriate one- or two-
tailed test (the former necessary only when the t-statistic is of the right sign), the
investigator then rejects the null hypothesis if this probability is sufficiently small.
But what do we mean by su fficiently small? The answer is by no means obvious,
and it depends upon the circumstances. It has be- come convention in social scientific
research to test one particular null hypothesisnamely, the hypothesis that the true
value of a co- efficient is zero. Under this hypothesis, in our notation above is equal
to zero, and hence the t-statistic is simply /s, the coefficient estimate divided by its
standard error. It is also convention to em- brace a significance level of .1o, . , o or
.o1that is, to inquire whether the t-statistic that the investigator has obtained, or one
even larger in absolute value, would arise more than 1o percent, percent, or 1 percent of
the time when the null hypothesis is correct. Where the answer to this question is no, the
null hypothesis is rejected and the coefficient in question is said to be statistically
significant. For example, if the parameter estimate that was obtained is far enough
from zero that an estimate of that magnitude, or one even further from zero, would
arise less than percent of the time, then the co- efficient is said to be significant at the
.o level.
The question whether the conventional social scientific signifi- cance tests are
appropriate when regression analysis is used for legal
:
See sources cited note 1 supra.
applications, particularly in litigation, is a difficult one that I will de- fer to the concluding
section of this lecture. I will simply assume for now that we are interested in the general
problem of testing some null hypothesis, and that we will reject it if the parameter
estimate obtained lies far enough out in one of the tails of the distribution from which
the estimate has been drawn. We leave open the ques- tion of what constitutes far
enough and simply seek to compute the probability under a one- or two-tailed test
of obtaining an estimate as far from the mean of the distribution as that generated by
the regression if the null hypothesis is true.
Most computerized regression packages report not only the pa- rameter estimate
itself ( in our notation), but also the standard error of each parameter estimate (s in
our notation). This value, coupled with the hypothesized true parameter value (
in our notation), can then be employed to generate the appropriate t- statistic for
any null hypothesis. Many regression packages also report a number called the t-
statistic, which is invariably based upon the conventional social scientific null
hypothesis that the true parameter value is zero. finally, some packages report the
probability that the t-statistic at issue could have been generated from a t-
distribution with the appropriate degrees of freedom under a one- or two-tailed test.
:6
Returning to tables 1 and , : all of this information is reported for each of the five
parameter estimatesthe standard error, the value of the t-statistic for the null hypothesis
that the true parameter value is zero, and the probability of getting a t-statistic that large
or larger in absolute value under a two-tailed test with | degrees of freedom. To interpret
this information, consider the estimated coefficient for the
:6
If the regression package does not report these probabilities, they can readily be found elsewhere. It
has become common practice to include in statistics and econometrics books tables of probabilities for t-
distributions with varying degrees of freedom. Knowing the degrees of freedom associated with a t-
statistic, there- fore, one can consult such a table to ascertain the probability of obtaining a t- statistic
as far from zero or farther as the one generated by the regression (the con- cept far from zero again
defined by either a one- or two-tailed test). As a point of reference, when the degrees of freedom are large
(say, o or more), then the .o significance level for a two-tailed test requires a t-statistic approximately
equal to
:.o.
gender dummy in table 1. The estimated coefficient of

1|,o.| has standard error of
1|o:.: and thus a t-statistic of

1|,o.|/1|o:. : =

1.o|q. The associated probability under a two-tailed test is reported
as .o. This means
that if the true value of the coefficient for the gender dummy were zero, a coeffi cient
greater than or equal to
1|,o.| in absolute value would nevertheless arise o percent of the time given the
degrees of freedom of the t-distribution from which the coefficient estimate is drawn. A
rejection of the null hypothesis on the basis of a parameter estimate equal to 1|,o.| or
greater in ab- solute value, therefore, will be erroneous three times out of ten when the
null hypothesis is true. By conventional social science standards, therefore, the
significance level here is too low to reject the null hy- pothesis, and the coefficient of the
gender dummy is not statistically significant. It is noteworthy that in this instance (in
contrast to any real world application), we know the true parameter value, namely

. :ooo o. Hence, if we employ a conventional two-tailed significance
test, we are led
erroneously to reject the hypothesis that gender dis- crimination is present.
As noted, we may regard the two-tailed test as inappropriate for the coefficient of the
gender dummy because we find the possibility of discrimination against men to be
implausible. It is a simple matter to construct an alternative one-tailed test: Table 1
indicates that a coefficient estimate of 1|,o.| or greater in absolute value will occur
o percent of the time if the true value of the coefficient is zero. Put differently, an
estimate of the gender dummy coefficient greater than or equal to 1|,o.| will arise 1
percent of the time, and an estimate
less than or equal to

1|,o.| will arise 1 percent of the time. It fol-
lows that if we are
only interested in the lower tail of the t-distribu- tion, rejection of the null hypothesis
(when it is true) will be erro- neous only 1 percent of the time if we require a
parameter estimate
of

1|,o.| or smaller. The one-tailed significance level is thus .1,
still below the
conventional thresholds for statistical significance.
:,
Using such significance levels,
therefore, we again are led to accept the null hypothesis, in this case erroneously.
:,
The result in this illustration is generalfor any t-statistic, the probability of rejecting the null
hypothesis erroneously under a one-tailed test will be exactly half that probability under a two-tailed test.
I off er this illustration not to suggest that there is anything wrong with
conventional significance tests, but simply to indicate how one reduces the chance of
erroneously rejecting the null hy- pothesis (call this a Type I error) only by increasing
the chance of erroneously accepting it (call this a Type II error). The conven- tional
significance tests implicitly give great weight to the impor- tance of avoiding Type I
errors, and less weight to the avoidance of Type II errors, by requiring a high degree of
confidence in the falsity of the null hypothesis before rejecting it. This seems perfectly
ap- propriate for most scientific applications, in which the researcher is justifiably asked
to bear a considerable burden of proof before the scientific community will accept that
the data establish an asserted causal relation. Whether the proponent of regression
evidence in a legal proceeding should bear that same burden of proof is a more subtle
issue.
r. Goodness of fit
Another common statistic associated with regression analysis is the R
2
. This has a
simple definitionit is equal to one minus the ratio of the sum of squared estimated
errors (the deviation of the ac- tual value of the dependent variable from the regression
line) to the sum of squared deviations about the mean of the dependent variable.
Intuitively, the sum of squared deviations about its mean is a mea- sure of the total
variation of the dependent variable. The sum of squared deviations about the
regression line is a measure of the ex- tent to which the regression fails to explain the
dependent variable (a measure of the noise). Hence, the R
2
statistic is a measure of the ex-
tent to which the total variation of the dependent variable is ex- plained by the
regression. It is not diffi cult to show that the R
2
statistic necessarily takes on a value
between zero and one.
:8
A high value of R
2
, suggesting that the regression model explains the
variation in the dependent variable well, is obviously important if one wishes to use the
model for predictive or forecasting purposes. It is considerably less important if one is
simply interested in particular parameter estimates (as, for example, if one is searching
for evidence of discrimination, as in our illustration, and thus
:8
See, e.g., E. Hanushek and J. Jackson, supra note , 1 at ,

8.
focused on the coefficient of the gender dummy). To be sure, a large unexplained
variation in the dependent variable will increase the standard error of the coefficients
in the model (which are a function of the estimated variance of the noise term), and
hence regressions with low values of R
2
will often (but by no means always) yield
parameter estimates with small t-statistics for any null hypothesis. Because this
consequence of a low R
2
will be reflected in the t- statistics, however, it does not
afford any reason to be concerned about a low R
2
per se.
As a quick illustration, turn back to tables 1 and . : Recall that the noise terms for the
data set from which the estimates in table 1 were generated were drawn from a
distribution with a standard deviation of , ooo, while for table : the noise terms were
drawn from a distri- bution with a standard deviation of , 1 ooo. The unexplained variation
in the earnings variable is likely to be greater in the first data set, therefore, and
indeed the R
2
statistics con firm that it is (.6|6 for table 1 and .q6| for table ). :
Likewise, because the estimated variance of the noise term is greater for the estimates in
table 1, we expect the coefficient estimates to have larger standard errors and smaller t-
statistics. This expectation is also borne out on inspection of the two tables. Variables
with coefficients that are statistically significant by conventional tests in table :,
therefore, such as the gender dummy, are not statistically significant in table 1.
In these illustrations, the value of R
2
simply reflects the amount of noise in the data,
and a low R
2
is not inconsistent with the mini- mum SSE criterion serving as an
unbiased, consistent, and efficient estimator because we know that the noise
terms were all independent draws from the same distribution with a zero mean. In
practice, however, a low value of R
2
may indicate that important and systematic factors
have been omitted from the regression model. This possibility raises again the concern
about omitted variables bias.
. Two Common Statistical Problems in Regression Analysis
Much of the typical econometrics course is devoted to what hap- pens when the
assumptions that are necessary to make the minimum SSE criterion unbiased, consistent,
and efficient do not hold. I can- not begin to provide a full sense of these issues in such a
brief lecture
and will simply illustrate two of the many complications that may arise, chosen because
they are both common and quite important.
. Omitted Variables
As noted, the omission from a regression of some variables that affect the dependent
variable may cause an omitted variables bias. The problem arises because any omitted
variable becomes part of the noise term, and the result may be a violation of the
assumption necessary for the minimum SSE criterion to be an unbiased estima- tor.
Recall that assumptionthat each noise term is drawn from a distribution with a
mean of zero. We noted that this assumption logically implies the absence of
correlation between the explanatory variables included in the regression and the
expected value of the noise term (because whatever the value of any explanatory
variable, the expected value of the noise term is always zero). Thus, suppose we start
with a properly speci fied model in which the noise term for every observation has an
expected value of zero. Now, omit one of the independent variables. If the effect of this
variable upon the de- pendent variable is not zero for each observation, the new noise
terms now come from distributions with nonzero means. One con- sequence is that the
estimate of the constant term will be biased (part of the estimated value for the
constant term is actually the mean effect of the omitted variable). Further, unless the
omitted variable is uncorrelated with the included ones, the coefficients of the included
ones will be biased because they now reflect not only an estimate of the effect of the
variable with which they are associated, but also partly the effects of the omitted
variable.
:q
To illustrate the omitted variables problem, I took the data on which the estimates
reported in table 1 are based, and reran the re- gression after omitting the schooling
variable. The results are shown in table :
:q
See J. Johnston, supra note 1, at 168

6q; E. Hanushek and J. Jackson, supra note 1, at 81

8:.
The bias is a function of two thingsthe true coe fficients of the excluded variables, and the correlation
within the data set between the included and the excluded variables.
Table . Omitted variable illustration
Variable True
value
Estimated
value
Standard
error
t-statis- tic Prob
(:-tail)
Constant ooo.o q8o6. |6.8 :.1o, .o|1
School 1ooo.o omitted
Aptitude o.o 1o,. :.6 |.1, .ooo
Experience oo.o :6.q 1o. :.|8, .o1,
Gendum

:ooo.o

:||. 1,,q.o

1., .1,6
R
2
= .|o8
You will note that the omission of the schooling variable lowers the R
2
of the
regression, which is not surprising given the original importance of the variable. It
also alters the coefficient estimates. The estimate for the constant term rises
considerably, because the mean effect of schooling on income is positive. It is not
surprising that the constant term is thus estimated to be greater than its true value. An
even more significant effect of the omission of schooling is on the coefficient estimate for
the aptitude variable, which increases dramatically from below its true value to well above
its true value and becomes highly significant. The reason is that the schooling variable is
highly correlated (positively) with aptitude in the data setthe correlation is .6 qand
because schooling has a positive effect on earnings. Hence, with the schooling variable
omitted, the aptitude coefficient is erroneously capturing some of the (positive) returns
to education as well as the returns to aptitude. The consequence is that the minimum
SSE criterion yields an upward biased estimate of the coefficient for aptitude, and in this
case the actual estimate is in- deed above the true value of that coefficient.
The effect on the other coefficients is more modest, though non- trivial. Notice, for
example, that the coefficient of gendum increases (in absolute value) significantly. This
is because schooling happens to be positively correlated with being male in my
fictitious data setwithout controlling for schooling, the apparent effect of gender is
exaggerated because females are somewhat less well educated on average.
The omitted variables problem is troublesome to investigators not simply because it
requires them to collect data on more variables to avoid it, but because the omitted
variables are often unobservable. Real world studies of gender discrimination are
perhaps a case in point. One can readily imagine that earnings depend on such factors as
innate ability and motivation, both of which may be unobservable to an investigator.
Omitted variables bias may then become some- thing that the investigator cannot avoid,
and an understanding of its consequences then becomes doubly important. For an
investigator concerned primarily with the coefficient of the gender dummy, it might be
argued, the omitted variables bias caused by the exclusion of innate ability and
motivation should be modest because the corre- lation in the sample between gender
and those omitted variables might plausibly be assumed to be small. Where the problem
appears likely to be serious, by contrast, the utility of conventional regression as an
investigative tool diminishes considerably.
o
I note in passing that the problem of including extraneous or ir- relevant variables is
less serious. Their expected coefficient is zero and the estimates of the other
coefficients are not biased, although the efficiency of the minimum SSE criterion is
lessened.
1
I also note in passing a problem that is closely related to the omitted variables
problem, termed errors in variables. In many re- gression studies, it is inevitable that
some explanatory variables will be measured with error. Such errors become part of the
noise term. Let us assume that in the absence of measurement error, the noise terms
obey the assumption needed for unbiasedness and consis- tencythey are all drawn
from a distribution with zero mean and are thus uncorrelated with the explanatory
variables. With measure- ment error, however, this assumption will no longer hold.
o
Econometricians have developed some more sophisticated regression tech- niques to deal with the
problem of unobservable variables, but these are not always satisfactory because of certain restrictive
assumptions than an investigator must make in using them. See, e.g., Griliches, Errors in Variables and
Other Unob- servables, | : Econometrica q,1 (1q,|). An accessible discussion of the omitted variables
problem and related issues may be found in P. Kennedy, supra note 1,, at
6q

,:.
1
Id.
Imagine, for concreteness, that earnings depend on education, experience, and so
on, as hypothesized earlier, and on innate ability as suggested above. Instead of
supposing that innate ability is an omitted variable, however, suppose that the aptitude
test score in- cluded in the regression is a proxy for innate ability. That is, we re- gard it
as an imperfect measure of ability, correlated with it but not perfectly. When the test
score underestimates ability, the noise term rises, and when the aptitude score
overestimates ability the noise term falls. The result is a negative correlation between the
noise term and the aptitude/ability variable. Put differently, if the noise term without
measurement error is drawn from a distribution with zero mean, then the noise term
including the measurement error is drawn from a distribution with a mean equal to the
magnitude of that error. The consequence once again is bias in the estimated
coefficients of the model.
:
r. Multicollinearity
The multicollinearity problem does not result in biased coeffi- cient estimates, but
does increase the standard error of the estimates and thus reduces the degree of
confidence that one can place in them. The difficulty arises when two independent
variables are closely correlated, creating a situation in which their effects are diffi- cult to
separate.
The following illustration will convey the essential intuition: Suppose that two
law school faculty members (call them Baird and Picker) regularly address alumni
luncheons, held partly for the pur- pose of stimulating alumni contributions. Assume
that each time one gives a luncheon speech, the other does too, and that the only
available datum on alumni contributions is aggregate monthly
:
One standard technique for addressing this problem is termed instrumental variables, which
replaces the tainted variable with another variable that is thought to be closely associated with it but also
thought uncorrelated with the disturbance term. For a variety of reasons, however, the instrumental
variables technique is not satisfactory in many cases, and the errors in variables problem is consequently
one of the most serious difficulties in the use of regression techniques. A discussion of the instrumental
variables technique and other possible responses to the errors in variables problem may be found in P.
Kennedy, supra note 1,, at 11

16 ; J. Johnston, supra note , 1 at :81

q1.
giving. We somehow know that each time both give a speech in a month, alumni
contributions rise by $1o,ooo. When they each give two speeches in a month,
contributions rise by $ , , :o ooo and so on.
Thus, by hypothesis, we know the joint effect of a speech by both Baird and Picker
($1 , o ooo), but nothing in the data permit us to as- certain their individual effects.
Perhaps each speech increases contri- butions by $,ooo, but it might be that one speaker
induces an extra
$1o,ooo in giving and the other none, or that one speaker induces an extra $o,ooo in
giving and the other reduces giving by $:o,ooo. In econometric parlance, the data on
speeches given by Baird and Picker are perfectly collinearthe correlation between
the number of speeches per month by Baird and the number per month by Picker is
1.o. An attempt to estimate the effect of the number of speeches given by each upon
contributions would fail and result in an error message from the computer (for reasons
that we need not detail, it would find itself trying to divide by zero).
The term multicollinearity usually refers to a problem in the data short of the
perfect collinearity in our illustration, but where changes in two variables are
nevertheless highly correlated to the point that it is difficult to separate their effects.
Because multi- collinearity does not go to any property of the noise term, the mini-
mum SSE criterion can still be unbiased, consistent, and e fficient. But the difficulty in
separating the e ffects of the two variables intro- duces greater uncertainty into the
estimator, manifest as an increase in the standard errors of the coefficients and a
reduction in their t- statistics.
One illustration of the effects of multicollinearity may already have been provided.
In our discussion of table 1, we noted that the coefficient estimate for the aptitude
variable was far below its true value. As it turns out, aptitude and schooling are highly
correlated in the data set, and this affords a plausible conjecture as to why the co-
efficient for the schooling variable is too high and that for aptitude insignificantly small
(some of the effects of aptitude in the sample are captured by the schooling coefficient).
To give another illustration, which incidentally allows us to in- troduce another use
of dummy variables, suppose that gender dis- crimination at our hypothetical firm
affects the earnings of women in two waysthrough an effect on the baseline earnings of
women as
before, and through an effect on the returns to education for women. In particular, recall
that in equation (1) both sexes earned $1,ooo per year of schooling. Suppose now that
males earn $ , 1 ooo, but females earn only $8oo. This effect can be captured
mathematically by an interaction term incorporating the gender dummy, so that
earnings are now determined in accordance with equation ( ): :
Earnings = 5000 + 1000 School + 50 Aptitude
+ 300 Experience

2000 Gendum

200 Gendum School ( ) :
Using the same hypothetical data for the explanatory variables as before, I produced
new values of earnings using equation (:) and (just for varietys sake) noise terms
drawn from a distribution with standard deviation of , : ooo. I then estimated a
regression from the new data set, including the variable Gendum School as an
additional explanatory variable. The results are in table |, where the variable Interact is
simply Gendum School.
Table |. Multicollinearity illustration
Variable True
value
Estimated
value
Standard
error
t-statis- tic Prob
(:-tail)
Constant ooo.o oo.: :1o.6 1.6| .1o8
School 1ooo.o q6:. 1||.o 6.6,q .ooo
Aptitude o.o 61.: 1:. |.q66 .ooo
Experience oo.o :88. 6., ,.861 .ooo
Gendum

:ooo.o

|:|., :q16.

1.| .1
Interact

:oo.o 1|.8 1q8.: .o, .q|1
R
2
= .qoq
Observe that, in contrast to table 1, the coe fficient for the gender dummy is now
higher than the real value by more than a factor of two. The coefficient for the
interaction term, by contrast, has the wrong sign and is close to zero. The other
parameter estimates are not too far off the mark.
The poor results for the coefficients of Gendum and Interact are almost certainly a
result of a severe multicollinearity problem. Note
that whenever Gendum = 0, Interact = 0 as well. Gendum is positive only when Interact
is positive. We would expect a high correlation between them, and indeed the
correlation coefficient is .q6. Under these circumstances, it is no surprise that the
regression cannot sepa- rate the effects of the two variables with any accuracy. The
estimated coefficient for Interact is insignificant by any plausible test, and the coefficient
for Gendum also has a large standard error that produces a rather poor t-statistic despite
the high absolute value of the coeffi- cient estimate.
Notwithstanding the considerable uncertainty introduced into the coefficient
estimates, however, it is plausible that the multi- collinearity problem here is not
disastrous for an investigator inter- ested in identifying the extent of gender
discrimination. The reason is that the estimate of the joint effects of the Gendum and
Interact may not be too far afieldone is inflated and one is understated, with the
errors to a great extent canceling each otherand as a legal matter an estimate of the
joint effect may be all that is needed. The caveat is that multicollinearity reduces the t-
statistics for both vari- ables, and might thereby lead the investigator to reject
the hypothesis that discrimination is present at all. To deal with the effects of
multicollinearity here, therefore, the investigator might simply wish to discount the low
t-statistics, or else to omit one of the two variables and recognize that the coeffi cient
estimate for the included variable will be biased and will include the effect of the
omitted variable.

In many instances, however, the investigator will not be satisfied with an estimate of
the joint effect of two variables, but needs to sep- arate them. Here, multicollinearity can
become highly problematic. There is no simple, acceptable solution for all cases, though
various options warrant consideration, all beyond the scope of this lecture.
|
. A final Note on the Law: Regression Analysis and the Burden of Proof
A key issues that one must confront whenever a regression study is introduced into
litigation is the question of how much weight to

It is important to recollect that this approach raises the problem of omitted variables bias for the
other variables as well.
|
See P. Kennedy, supra note , 1, at 1|6

6.
give it. I hope that the illustrations in this lecture afford some basis for optimism that
such studies can be helpful, while also suggesting considerable basis for caution in their
use.
I return now to an issue deferred earlier in the discussion of hy- pothesis testingthe
relationship between the statistical significance test and the burden of proof. Suppose, for
example, that to establish liability for wage discrimination on the basis of gender under
Title VII, a plainti ff need simply show by a preponderance of the evidence that women
employed by the defendant suffer some measure of dis- crimination.

With reference to
our first illustration, we might say that the required showing on liability is that, by a
preponderance of the evidence, the coefficient of the gender dummy is negative.
Unfortunately, there is no simple relationship between this bur- den of proof and the
statistical significance test. At one extreme, if we imagine that the parameter estimate in
the regression study is the only information we have about the presence or absence of
discrimi- nation, one might argue that liability is established by a preponder- ance of
the evidence if the estimated coefficient for the gender dummy is negative regardless
of its statistical significance or standard error. The rationale would be that the negative
estimate, however subject to uncertainty, is unbiased and is the best evidence we have.
But this is much too simplistic. Very rarely is the regression es- timate the only
information available, and when the standard errors are high the estimate may be among
the least reliable information available. Further, regression analysis is subject to
considerable ma- nipulation. It is not obvious precisely which variables should be in-
cluded in a model, or what proxies to use for included variables that cannot be measured
precisely. There is considerable room for exper- imentation, and this experimentation
can become data mining, whereby an investigator tries numerous regression
specifications until the desired result appears. An advocate quite naturally may have a
tendency to present only those estimates that support the clients position. Hence, if
the best result that an advocate can present contains high standard errors and low
statistical significance, it is often plausible to suppose that numerous even less
impressive

See, e.g., Texas Department of Community Affairs v. Burdine, |o U.S. :|8

(1q81).
results remain hidden, and conceivably shielded from discovery by the work product
doctrine.
6
For these reasons, those who use regression analysis in litigation tend to report
results that satisfy the conventional significance testsoften the -percent
significance leveland to suppose that less signi ficant results are not terribly
interesting.
,
Before most ex- perts would feel comfortable asserting that gender
discrimination has been established by a study such as that in our illustration, therefore,
they likely would require that the coe fficient estimate for the gender dummy be negative
and statistically significant. Even then, they would anticipate a vigorous cross-
examination based on a number of matters, many suggested by the discussion above.
Still more difficult issues arise when an exact parameter estimate is needed for some
purpose, such as for computing damages. The fact that the parameter is statistically
significant simply means that by conventional tests, one can reject the hypothesis that its
true value is zero. But there are surely many other hypotheses about the pa- rameter
value that cannot be rejected, and indeed the likelihood that regression will produce a
perfectly accurate estimate of any parameter is negligible. About the only guidance
that can be given from a statistical standpoint is the obviousparameter estimates
with proportionally low standard errors are less likely to be wide of the mark than others.
Ultimately, therefore, statistics itself does not say how much weight a regression
study ought be given, or whether it is reasonable to use a particular parameter estimate
for some legal purpose or other. These assessments are inevitably entrusted to triers
of fact, whose judgments on the matter if well informed are likely as good as those of
anyone else.
6
I will not digress on the rules of discovery here. In practice, the raw data may be discoverable,
for example, while the experts undisclosed analysis of the data may not be.
,
See the discussion in Fisher, Statisticians, Econometricians and Adversary
Proceedings, 81 J. Am. Stat. Assn. :,, ( ). 1q86